PROJECT SUMMARY Studies using human tumor samples have shown MDM2 gene amplification in a variety of human tumor types, most notably, sarcomas.In human tumors, the MDM2 gene is found as part of an amplicon derived from chromosome 12q.14-15. This amplicon frequently also contains the CDK4 and HMGA2 genes. Studies in liposarcoma as well as other tumor types have indicated that most of the genes located between CDK4, HMGA2, and MDM2 often are not found to be amplified. This suggests that they are on three distinct amplicons or there is no selective pressure to retain genes that are found between them. To date, there have only been limited mouse models that address the role of Mdm2 in tumorigenesis. None of these have attempted to specifically model human disease. Inhibitors of the interaction of Mdm2 with p53 have been extensively studied for their potential utility as targeted therapies but such studied have been hampered by the lack of suitable preclinical models. In the studies proposed here, a set of mouse models will be generated that overexpress Mdm2, Cdk4, or Hmga2. These will be used to phenocopy specific human cancers that are associated with MDM2 gene amplification and the 12q amplicon, namely liposarcoma and osteosarcoma. The generation of such mouse models will facilitate elucidating the underlying molecular basis for the oncogenic activity of Mdm2 and the 12q amplicon, as well as provide utility in preclinical testing of novel targeted therapies. It should be emphasized that there is an urgent need for mouse models that mimic Mdm2- and 12q amplicon-driven tumorigenesis in humans for two compelling reasons. First, such models are needed to fully determine the molecular basis for oncogenic activity of Mdm2 and other 12q amplicon genes in order to further design and implement suitable target therapies. Second, such preclinical models are needed to validate already established approaches involving chemical inhibition of the binding of Mdm2 to p53.